w210 Capstone.
====


- No need to code from scratch:
- 14 weeks scope.
- Credit anythings that's use.
- Public data is actually best and preferred
- Form need to be field out.
- Create a web platform access to everyone.
- Access to the data.
- Focus on MVP.

This is what we can do this.
Together we can do this.

Ideas Generations:


Proposal 1: A platform to Help Tuberculosis clinician/researchers accross the world to predict TB treatment outcome using ML model with Explaiable AI using NIH Tuberculusis Data.

Proposal 2: Classificy Public social media account aimed proned to be conducting Human trafficing. Use Social media such as FACEBOOK, TWITTER public data/ post to identifie post which target that Specific age group or a particular country to predict human trafficing.

Proposal 3: use Machine Learning, to recognize patterns of false accounts for conviction.

# Proposal 1
- Pro* Problem Clarity:

* Potential Impact:
*
* Feasibility
*
* Success evaluation.

help you understand and interpret predictions made by your machine learning models
*
Help TB researchers accross the world to predict TB treatment outcome using ML model with Explaiable AI.

My Team manage the platform, API, request approval and only a data use agreement.

We made a site available with details on how to use the data

The Tuberculosis Data Exploration Portal (TB DEPOT) has been created for working with cohorts that one can analyze across four types of data: clinical, genomic sequences, CT scans, and X-ray iamge data data. The TB DEPOT enables basic querying and building of cohorts, and includes some basic exploratory analytics tools.

We can beggin initial work while we await approval of your submitted Data Use Agreement.

The data is publicly available.
I do have Access to subject matter expert, researcher at the NIH etc...
Feasinility

https://datasharing.tbportals.niaid.nih.gov/#/data-download-options

https://datasharing.tbportals.niaid.nih.gov/#/home?view=tb-depot

# Proposal 2
## Version A
Help Law unforcement predict, and traffic
Human trafficing Prediction using public data.
Use hotel booking data,
geo-referencing, GPS-based mobile Web application for members for law enforcement to detect and predict areas trafficking victims and activate immediate direct assistance.

Human traffiking increase during event like superblow
Use call 911 call data

Help predict zone more like to victim of human trafficing, flag social organization and assist with care.

## Version B
Use Social media such as FACEBOOK, TWITTER public data/ post to identifie post which target that Specific age group or a particular country to predict human trafficing. Using classification through NLP. Predict Crimimal and note criminal account.
NER will be use to ID certain keywor related to threat, location etc..
Build platform,To itenfied accounts with higher chance of being involve in related activity.

- Potential impact

- Feasibility:
Use facebook graph API. Available and can be run using python.

# Proposal 3:
-


------------------

Help clinicians and researchers predict , Tuberculosis treatment outcomes of different cohorts using ML models with explaiable AI leveraging NIH Tuberculosis Data.

Prediction of human trafficking and harmul social media accounts using public data and posts such as Facebook and Twitter to identify harmful posts that target a specific age group or particular country.

Use Machine Learning, to recognize patterns of false accounts, accurate and inaccurate eyewitnesses statement, for crime conviction during trial.

Potential capstone:

Proposal 2: Decode Nutrition & Ingredients labels - Read the ingredients labels without any extra-long chemical names! Understand the ingredients of the food products like a pro and chose wisely.

Proposal 2: AI world without traffic signals A tool to autonomously control traffic signals to minimize or potentially eliminate vehicle idle time at intersections

--------------------
Rajiv Bone deformoties

Satish Kumar.

----------------------

## Ethics Weeks
-